<p>Waveguides are used to transmit electromagnetic signals.Their geometry is typically long and slender their particularshape can be used in the design of computational methods. Onlyspecial modes are transmitted and eigenvalue and eigenvectoranalysis becomes important.</p><p>We develop a .nite-element code for solving theelectromagnetic .eld problem in closed waveguides .lled withvarious materials. By discretizing the cross-section of thewaveguide into a number of triangles, an eigenvalue problem isderived. A general program based on Arnoldi’s method andARPACK has been written using node and edge elements toapproximate the .eld. A serious problem in the FEM was theoccurrence of spurious solution, that was due to impropermodeling of the null space of the curl operator. Therefore edgeelements has been chosen to remove non physical spurioussolutions that arises.</p><p>Numerical examples are given for homogeneous andinhomogeneous waveguides, in the homogeneous case the resultsare compared to analytical solutions and the right order ofconvergence is achieved. For the more complicated inhomogeneouswaveguides with and without striplines, comparison has beendone with results found in literature together with gridconvergence studies.</p><p>The code has been implemented to be used in an industrialenvironment, together with full 3-D time and frequency domainsolvers. The2-D simulations has been used as input for full3-D time domain simulations, and the results have been comparedto what an analytical input would give.</p>

No description available.

No description available.

No description available.

"Transmission loss needs to be considered in the design of telecommunication systems. If telecommunication systems have high transmission loss, the signals lose too much of their strength, which results in poor reception in television networks and lost calls in cellular networks. Total transmission loss, in the MHz-GHz range, has several different loss components, some of which are poorly characterized. Conductor loss is the largest loss component and the most difficult to predict. It is known that the conductor geometry or features influences the conductor loss. However, current numerical, analytical and empirical tools do not accurately predict this loss component, and there is little experimental data available to explain and show the impact of these conductor geometries. The conductor shape is heavily influenced by the ceramic substrate surface roughness, and this is especially true for printed circuit boards fabricated with thick-film technology. The two conductor features of interest are the conductor-edge angle and conductor-ceramic interface. For thick-film circuits, the edge of the conductor does not have a square cross section but has a tapered shape or angle. The conductor-ceramic interface is also rough at the micron scale. Since the current density is concentrated at the extremities of the conductor then these features, conductor-ceramic interface and conductor edges, can potentially have a large impact on conductor loss. For this study, the surfaces of ceramic substrates were subjected to different surface finishes that resulted in distinctly different surface characteristics. This in turn resulted in a range of conductor-ceramic interfaces and conductor-edge angle geometries. The impact of the conductor-edge angle and conductor-ceramic interface features on conductor loss was measured over a range of frequencies and conductor conductivities to ascertain the level of their contribution. It was shown quantitatively that the conductor-edge angle was significantly altered by the surface roughness and heavily influenced the conductor loss result. The consensus for decades has been that greater surface roughness causes the ceramic-conductor interface geometry to have a greater impact on conductor loss, increasing the conductor loss. However, this study has shown that greater surface roughness also causes the conductor-edge angle feature to have a smaller or reduced impact on conductor loss, improving the conductor loss result - this has not been considered previously. Focusing on only one of these features can give an anomalous loss prediction; both features need to be considered for the calculation of conductor loss for thick-film applications. The low frequency loss results are as expected but the high frequency (greater than 5GHz) results depend on edge angle and therefore thick-film paste viscosity, and substrate surface roughness. "

El objetivo del presente trabjo de título es entender el funcionamiento de una red GPRS, identificando cada uno de sus componentes e interfaces, para luego proponer un modelo teórico simple que permita dimensionar la capacidad y caracterizar la calidad de una red GPRS. El trabajo se enmarca en la iniciativa de la empresa de telefonía celular Movistar Chile, en realizar este en conjunto con el Departamento de Ingeniería Eléctrica de la Universidad de Chile, para sentar las bases de un intercambio de conocimiento tecnológico.

High power, broadband light sources emitting in the 2-2.5 um wavelength range important for optical sensing of important biomolecules in aqueous solutions such as glucose. Here we demonstrate and analyze superluminescent diodes with output centered at 2.4 ums (range ~2.2-2.5 ums) from GaInAsSb/AlGaAsSb quantum wells in a separate confinement structure. Quasi-continuous wave output of 1 mW is achieved at room temperature for 40m x 2mm devices. Superluminescence is evidenced in superlinear growth, spectral narrowing, and angular narrowing of light output with increasing current injection. The output is analyzed and modeled with semiconductor rate equations, and by varying parameters, potential routes for future improvements are explored, such as additional Auger suppression and photonic mode engineering.

No description available.

In this thesis, we investigate the contraction reconstruction conjecture. It states that all simple graphs with at least four edges are reconstructible, that is they are uniquely determined from their collection of single edge contraction minors, called the deck. Similar questions have been studied in the past, the vertex reconstruction conjecture being the most famous. There are usually two steps to show that a class of graph is reconstructible. The first one is to show that the class is recognizable, meaning that it is possible to determine if a graph G belongs to that class by looking at its deck. In order to recognize some classes of graphs, we show that a wide range of graph properties are reconstructible. We investigate the connectivity of graphs, which is useful to recognize disconnected, separable, and 2-connected graphs. We also show that the number of cycles of various lengths, the degree sequence, the number of spanning trees, the planarity, the presence of cliques of various sizes, and the diameter are reconstructible. Knowing the lengths of cycles allows us to recognize the class of bipartite graphs, while knowing the degree sequence allows us to recognize regular graphs. The second step in showing that a class of graph is reconstructible is called weak reconstruction. We say that a class of graph is weakly reconstructible if no two graphs in that class share the same deck. A class of graphs that is both weakly reconstructible and recognizable is reconstructible. In this thesis, we show that disconnected graphs, bipartite graphs, most separable graphs and most 2-edge connected graphs are reconstructible. We also show that distance regular graphs and some cubic graphs are reconstructible. We quickly delve into the theory of probabilities to give a proof that almost all graphs are reconstructible. Finally, the relation between edge contraction and graph automorphisms is studied. We study the automorphism group of a graph in relation to those of its cards. We also study the concept of contraction pseudo-similarity. Two edges are contraction pseudo-similar if they are not similar, but their contractions yield isomorphic graphs. We completely characterize the graphs that contain contraction pseudo-similar edges.

The blurry, undefined edge of a city “is a philosophical region, where city and landscape overlap, existing without choice or expectation.” 1 This urban condition hardly ever influences the city’s core, leaving the periphery isolated, autonomous, and untapped. In this blurry zone, architectural opportunities present themselves. In New Orleans, a small Vietnamese community called Versailles lives in this undefined realm at the eastern boundary of the city. Its residents are removed from the core, disconnected by sheer distance, and wedged between ecology and industry. Typically, “that which literally lies on the margins often gets figuratively marginalized,” 2 but this community is resilient, taking full advantage of its overlap with the natural environment. A product of a forced migration in the wake of Saigon’s fall in 1975, this community drew on local parallels to Vietnamese climate and fishing commerce, allowing newcomers to morph this neighborhood into their own. Yet the scars of heavy industry still plague Versailles, which has endured a fraught history with nearby landfills. This thesis seeks to create productive frictions between community, ecology, and industry through exploring a blended topography where construction and landscape are commingled. This will allow for the architecture to be both anchored and stretched. The Eco-Line - part composting and gardening hub, part cultural and economic incubator, part ecological and recreational park - weaves a new edge of experience on New Orleans’s periphery. As opposed to forcing a connective intervention inward on New Orleans, the Eco-Line hugs the easternmost edge of Versailles, exploring a new urban condition that blends with the natural environment. / 0 / SPK / specialcollections@tulane.edu